Heterogeneity

Heterogeneity in a meta-analysis measures how much individual study results vary beyond what would be expected from sampling error alone. High heterogeneity (I² > 75%) suggests that studies are measuring different underlying effects, potentially due to differences in populations, dosing protocols, supplement forms, or outcome measures. When heterogeneity is high, the pooled estimate may not be meaningful, and subgroup analyses or meta-regression should explore sources of variation. The Q statistic tests for the presence of heterogeneity, while I² quantifies its magnitude.